|
|
Deep Learning for Image Recognition
Munzar, Milan ; Kolář, Martin (referee) ; Hradiš, Michal (advisor)
Neural networks are one of the state-of-the-art models for machine learning today. One may found them in autonomous robot systems, object and speech recognition, prediction and many others AI tasks. The thesis describes this model and its extension which is used in an object recognition. Then explains an application of a convolutional neural networks(CNNs) in an image recognition on Caltech101 and Cifar10 datasets. Using this exemplar application, the thesis discusses and measures efficiency of techniques used in CNNs. Results show that the convolutional networks without advanced extensions are able to reach a 80\% recognition accuracy on Cifar-10 and a 37\% accuracy on Caltech101.
|
|
|
Using machine vision for robot guidance
Gábik, Jaroslav ; Štěpánek, Vojtěch (referee) ; Vetiška, Jan (advisor)
With the development of machine vision technologies, new applications that can increase production, versatility or simplicity of production systems are widely spread. This thesis deals with usage of machine vision for robot guidance. The task consists of creating a technique and its practical realization, where the proposed assumptions are verified. The main objective is to determine 3D position and orientation of a sheet metal part or subassembly of the body-in-white, which is lying within the reach of an industrial robot with respect to its base coordinate system. The proposed method is suitable for several types and dimensions of components, which meet certain requirements. Targeting the component is carried out by scanning significant points on the component with the help of the 3D scanner attached to the robot flange. Afterwards, gained data are processed in a designed programme. The theoretical part is focused on research in the field of machine vision, accuracy of industrial robots, compensation of their errors and manipulation and assembly of the sheet metal parts in automotive. Finally, an evaluation and recommendations for practice are provided.
|
|
|
Automatic Image Labelling
Lukáč, Michal ; Řezníček, Ivo (referee) ; Hradiš, Michal (advisor)
This thesis focuses on automatic image labelling to semantic categories. It describes the theory of classif cation and local features detection. It explains fundamental machine learning models used for image tagging, and how such models can be learned with Gradient descent. It propose solution with hierarchy for ImageNet and tagging images with attributes. MapReduce computing model is considered for learning on big data sets. In the last part it is described implementation, experimental and test results.
|
|
|
Convolutional Neural Networks
Lietavcová, Zuzana ; Zbořil, František (referee) ; Zbořil, František (advisor)
This thesis deals with convolutional neural networks. It is a kind of deep neural networks that are presently widely used mainly for image recognition and natural language processing. The thesis describes specifics of convolutional neural networks in comparison with traditional neural networks and is focused on inner computations in the process of learning. Convolutional neural networks typically consist of a different types of layers of neurons and the core part of this thesis is to demonstrate computations of individual types of layers. Learning demonstrating program of a simple convolutional network was designed and implemented using own implementation of neural network. Validity of the implementation was tested by training models for solving a classification task. Experiments with different types of architectures were conducted and their performance was compared.
|
|
|
Recognition of Driving Lane Borders in Video from On Board Camera
Letovanec, Lukáš ; Bartl, Vojtěch (referee) ; Herout, Adam (advisor)
This thesis is dedicated to the issue of driving lane borders recognition in frames of an onboard camera. In this thesis, an architecture of a deep convolutional neural network is introduced, by means of which the said problem is dealt with. The net was trained on a large dataset using gradient descent algorithm. The trained model has demonstrated the ability to recognize borders of a driving lane well in different situations and conditions. The result of the thesis confirms that deep convolutional neural networks are a suitable tool for driving lane borders recognition.
|
|
|
Linear Logistic Regression Demo
Bak, Adam ; Kesiraju, Santosh (referee) ; Beneš, Karel (advisor)
This bachelor's thesis deals with the machine learning model logistic regression.The aim is to closely inspect and analyze the workings of this model for classification, in order to be able to provide a learning tool in the form of demonstrative application. All of the mathematical formulae, logistic sigmoid, cross entropy error function and gradient are derived and explained in a concise manner. This thesis also provides some insight into the form of the cross entropy error function in the case of linear logistic regression.
|
|
|
Neural Networks for Autonomous Car Driving
Dopita, Marek ; Hradiš, Michal (referee) ; Smrž, Pavel (advisor)
In this work, the principles of neural networks are introduced with a focus on autonomous vehicles. Based on this information, a proposal for the implementation of a system is created, which allows to drive a car without a driver. It builds on tools that allow easy creation and testing of autonomous vehicles. It is CARLA simulator and ranking.The proposal divides vehicle routes into three different situations. Each situation requires the use of different sensors, so a specific autonomous agent is created that is able to recognize the situation and switch between different neural network designs. Each such network is specific in its inputs and is taught in a specific situation.Programs are created that are able to easily collect a data set using the CARLA Leaderboard. Then, a way is introduced to how the collected data can be divided into categories so that each category can be used to learn its neural network.
|
|
|
Tram Detection in Video by Neural Network
Golda, Vojtěch ; Špaňhel, Jakub (referee) ; Dyk, Tomáš (advisor)
This paper deals with tram detection in video using convolutional neural networks. The basic principles of their function are described. A number of distinct architectures are trained. The usefulness of the resulting models is subsequently compared. The output of this paper is a program capable of detecting trams in video.
|
|
|
Tram Detection in Video by Neural Network
Golda, Vojtěch ; Špaňhel, Jakub (referee) ; Dyk, Tomáš (advisor)
This paper deals with tram detection in video using convolutional neural networks. The basic principles of their function are described. A number of distinct architectures are trained. The usefulness of the resulting models is subsequently compared. The output of this paper is a program capable of detecting trams in video.
|
|
|
Optimizations and applications of non-linear spectral unmixing in flow cytometry
Nemec, Matěj ; Musil, Jan (advisor) ; Stuchlý, Jan (referee)
Recent advances in flow cytometry techniques enable high-throughput single-cell experiments with extensive marker sets. In order to leverage this technology the measured signal must be unmixed to recover interpretable re- sults. Current approaches to unmixing typically leverage linear deconvolution algorithms such as fitting by ordinary least squares method, that tend have issues dealing with various noise sources inherit to the data collection pro- cess. This thesis evaluates the performance of a novel non-linear approach of unmixing called nougad. For the evaluation, we have generated realistic arti- ficial data with known ground truth for testing, implemented multi-threaded version of nougad and tuned its hyperparameters using Bayesian optimiza- tion, and collected several performance metrics of nougad and the other algorithms on the testing datasets. The results show that nougad is able to outperform the tested linear algorithms making this non-linear method more suitable for practical applications and a good candidate for further refinement and optimization efforts. 1
|
guest ::
